It can add a common measure
Caltech can read an SAT or ACT beside students from very different high schools. The score may clarify academic preparation when the local transcript is unfamiliar, but its meaning still depends on course access and performance.
It cannot replace curriculum
A section in Bucket B can still support a successful application; repeating a nearly perfect test may be less useful than strengthening STEM evidence. Admissions readers can see whether the student pursued the strongest appropriate sequence; a composite does not manufacture missing preparation.
It can reveal a section imbalance
SAT 780–800 or ACT sections 35–36 form Bucket A; SAT 750–770 or ACT 33–34 form Bucket B. A section-level plan is therefore more useful than repeatedly chasing a round total without diagnosing the limiting skill.
It cannot predict admission
Caltech evaluates individual sections in score buckets rather than treating one composite as the decision. The cohort contains admitted students, not a mechanical rule. Students below, inside and above a range can receive different decisions because the application is broader.
It can guide timing
Applicants self-report and only matriculating students submit official verification. The useful final sitting must leave enough time for the route Caltech accepts and for the applicant to confirm that the checklist is complete.
It cannot freeze the policy
For Caltech, current required-testing implementation describes the verified scope. Testing policies, optional sections and reporting instructions can change; the application term must remain attached to every operational decision.